Get updates on new articles, webinars and other opportunities:

Three Testing Strategies For Sophomore Conversion Testers

by Celine Roque

Change the color of your call-to-action button.  Test a new headline.  Swap this hero image for that hero image.

Been there, done that.

Try these three conversion testing methods for Conversion Testing sophomores.

#1 – Blow it up, start from scratch

Instead of pursuing split tests on specific site elements – such as the headline or call-to-action button – 37signals decided to test two fundamentally different versions of their Highrise homepage. The rationale behind this was that they needed to destroy their assumptions about what may or may not work.

Almost everything was different about each version.

This is an image of the Original Design vs. the Person Design:

  • The Original Design had several smaller customer photos, while the Person Design had one large background photo of just a single customer.
  • There were several customer testimonials in the Original Design, while the Person Design only had a single quote from the featured customer.
  • Only a handful of benefits are listed on the Person Design, while several features and benefits are outlined in the Original Design.
  • The person design has only one prominent call-to-action: “See Highrise Plans and Pricing”. The Original Design had a navigation menu, as well as options to view more testimonials and features.

The result of this difference was that the Person Design led to a 102.5% increase in paid sign-ups.

Key Takeaway: Rather than just testing small elements of your landing page, try to test radically different versions of the page.

#2 – Remove something “essential”

We all want our websites to have clear navigation – this makes it easier for visitors to browse our websites.

But does this make sense for all situations?

According to this test conducted for Yuppiechef, eliminating your navigation menu may help increase conversions. In their case, the increase was at 100%.

Why did this work? Having the navigation menu on their landing page distracted visitors from the primary call to action.

With too many options for what to do next, the visitor may be overwhelmed. Or they may attempt to manually find the information they need. Letting your users roam may be fine for most of your information pages or your blog, but sometimes it’s best to carefully direct these users to perform specific actions.

While this doesn’t mean abandoning notions of easier site navigation, it’s important to minimize other potential visitor actions on landing pages where your conversions matter the most.

Key Takeaway: Some items that may seem necessary, such as your navigation menu, could be getting in the way of higher conversions.

#3 – Test the audience, not the page

It’s simple to just test page elements, but how about testing which audience segments are most profitable? This is what Art Beads did during their holiday promo. Unlike the previous examples, this test was conducted via an email marketing campaign rather than a website.

While Art Beads could have sent a blast promo emails to their entire list, doing so made no sense for their business in the long run. A large scale campaign like this meant that their customers might expect discounts all the time.

Their solution? Find and test more profitable segments from their existing customer list. They identified that segment as current email subscribers who made only 1 purchase in the past 18 months – yet that purchase had to rank among the top 25% in value within the same period.

They ended up with a conversion rate 208% higher than their large-scale email blasts.

Key Takeaway: Increasing conversions doesn’t necessarily mean changing an element of your business or website. You can also try changing the customer segment you’re selling to.

One Comment


Get updates on new articles, webinars and other opportunities:

Celine Roque

Celine Roque writes about business, entrepreneurship, and the web. You can follow her on Google+.


Comment Policy

Please join the conversation! We like long and thoughtful communication.
Abrupt comments and gibberish will not be approved. Please, only use your real name, not your business name or keywords. We rarely allow links in your comment.
Finally, please use your favorite personal social media profile for the website field.


Your email address will not be published.

  1. Franz Keim says:
    October 23, 2012 at 3:36 am

    Nice article, Celine! For me one of the most important takeaways from testing is to be radical in what you test when you want to see big results. So I would fully subscribe to your first two points. Personally, I could never replicate one of the wonder stories where minor tweaks resulted in big lifts.

Show Me My Heatmap

I’m just @crazyegg’ing everything these days.

Kyle Mitchell